Northern Cascadia Research Articles

Archaeological Evidence for Village Abandonment Associated with Late Holocene Earthquakes at the Northern Cascadia Subduction Zone

Geologic evidence suggests that great (magnitude 8 or larger) earthquakes, or series of such earthquakes, occurred six times in the past 3000 yr at the northern Cascadia subduction zone. The archaeological record, and native oral traditions, demonstrate that native villages along the adjacent coasts of southern British Columbia and Washington State were occasionally abandoned in the late Holocene as a result of these earthquakes and associated tsunamis. We infer the temporal pattern of village occupation and abandonment from midden stratigraphy and from an activity index based on the probability distributions of radiocarbon ages at 30 archaeological sites in three regions of northern Cascadia. Deposits of probable tsunami origin are interbedded with, or bound, cultural strata at several sites. Earthquakes probably predate hiatuses in occupation, or periods of low inferred human activity, at many sites. The strongest correlation between earthquake incidence and site abandonment occurs in the Nootka Sound region. Effects of tsunamis vary with village location, coastal morphology, and late Holocene sea-level history.

Crustal structure of the Cascadia subduction zone, southwestern British Columbia, from potential field and seismic studies

The northern Cascadia subduction zone is a region of convergence between the oceanic Explorer and northern Juan de Fuca plates and the continental North American plate. Potential field and new seismic reflection data coupled with previous seismic results and geology enable derivation of a series of density – magnetic susceptibility cross sections and a block density model from the ocean basin to the volcanic arc from 2.5- and 3-dimensional interpretations. The lateral extent and thickness of the accreted wedge vary significantly along the zone. The narrow, metasedimentary Pacific Rim terrane lies immediately west of Wrangellia and extends the length of Vancouver Island, consistent with its emplacement by strike-slip faulting following the accretion of Wrangellia. The ophiolitic Crescent terrane is a narrow slice lying seaward of the Pacific Rim terrane but not extending northward beyond the Juan de Fuca plate. In this region, the Crescent terrane was emplaced in a strike-slip or obliquely convergent style during the latter stages of emplacement of Pacific Rim terrane. Below the accreted terranes, the Explorer plate is shallower than Juan de Fuca plate, resulting in a thinner crust. High-density lower crustal material lies beneath the western edge of Vancouver Island, supporting interpretations of wide-scale underplating of Wrangellia. The shape of the boundary region between Wrangellia and the Coast belt to the east varies along strike and may be controlled by properties of preexisting plutonic rocks. The low-density Coast belt plutons extend throughout most of the crust and are underlain by a lowermost crustal high-density layer, which may be indicative of fractionation accompanying magma generation.

Seismic reflection constraints on imbrication and underplating of the northern Cascadia convergent margin

An interpretation of the deep structure of the continental shelf offshore southern Vancouver Island, subject to constraints from other geophysical data, is derived by combining seismic reflection profiles shot in 1989 with those from an earlier 1985 survey. Accretionary wedge sediments, which extend landward beneath the volcanic Crescent terrane, comprise two primary units, both of which have shortened through duplex formation. The maximum thickness of the Crescent terrane, 6–8 km, occurs just seaward of its contact with the inboard, largely metasedimentary Pacific Rim terrane. The E region of reflectivity, first detected dipping landward beneath Vancouver Island, is regionally extensive, being observed on all the seismic profiles. The E reflectivity thins seaward and splits into two or more strands that probably link into major faults within the accreted sedimentary wedge. Reflections from the interplate décollement beneath the outer continental shelf separate from the downgoing plate, continue into the deepest level of the E reflectivity, and are interpreted to represent a single décollement surface above which imbrication of accreted units occurred. It is proposed that at the southern end of Vancouver Island the E reflections represent mainly underthrust sediments above a former subduction décollement, both of which were incorporated into the overlying continent when the subduction thrust stepped down into the descending oceanic plate. This change in depth of the subduction thrust underplated one or more mafic units to the continent. The reflection from the top of the subducting Juan de Fuca plate appears to be around 5 km shallower farther north along the margin, indicating that the underplated region could be confined to the embayment in the Cascadia subduction zone.

Seismic velocity increase and deep‐sea gas hydrate concentration above a bottom‐simulating reflector on the northern Cascadia continental slope

The amount of gas hydrates in the accretionary wedge sediments of the northern Cascadia subduction zone off Vancouver Island has been estimated from multichannel seismic (MCS) and Ocean Drilling Program (ODP) data. Detailed semblance velocity analyses and full waveform inversion of MCS data, combined with previously published ODP Sites 889/890 sonic log and vertical seismic profile (VSP) data, show that sediment velocities increase downward more rapidly than the no‐hydrate/no‐gas reference profile from about 1500 m/s near the seafloor to a maximum of 1900 m/s just above the bottom‐simulating reflector (BSR) at a subbottom depth of 224 m. Immediately below the BSR, the MCS velocities drop to ∼1700 m/s. A low velocity of ∼1500 m/s from the VSP data probably represents a thin layer, ∼20 m, containing free gas. The difference between the reference and observed velocities is used to estimate hydrate concentration, which reaches a maximum of 20–30% of the pore space above the BSR. A simple interpretation of the drill core chlorinity dilution data at the ODP site yields a similar hydrate concentration of 35%. The estimated hydrate concentration with subbottom depth represents about 7 m3 of hydrate per square meter of seafloor. The total methane gas at STP, including hydrate above the BSR and the small amount of free gas below, is about 800 m3 per square meter of seafloor, or 200 TCF (trillion cubic feet) on the Vancouver Island continental slope. Application of the same method in the Blake‐Bahama region of the eastern U.S. margin indicates that the velocity enhancement and inferred hydrate concentration are very similar.

Case for very low coupling stress on the Cascadia Ssubduction Fault

A fundamental problem in plate tectonics is the shear strength of major plate boundary faults. This translates to the question whether the generally observed small earthquake stress drops of 3–10 MPa on major faults release most of the accumulated stress or only a small fraction of it. There is strong evidence that the San Andreas fault, a major transform plate boundary, is weak (<20 MPa shear resistance). It is not yet clear whether subduction thrust faults are also weak. We present two types of evidence from the northern Cascadia subduction zone that indicate very low coupling shear stress on that plate interface and hence very low strength of the subduction thrust fault, comparable to that estimated for the San Andreas fault. First, the well‐defined surface heat flow and heat generation allow negligible frictional heating on the plate interface. The average shear stress on the fault must thus be very low over a time scale of a few million years. Second, focal mechanism solutions for small crustal earthquakes in the southern Vancouver Island area indicate that the horizontal stress in the direction of plate convergence has a similar magnitude to the vertical stress. This inferred stress state requires the present tectonic stress coupled across the subduction thrust fault to be very low. One explanation for the weakness of the fault is the presence of near‐lithostatic pore fluid pressure in the region of the fault zone for which there is independent evidence. The conclusion of a weak subduction thrust fault does not conflict with geodetic observations of contemporary surface deformation which indicate that the fault is currently locked, accumulating strain energy toward a future great earthquake. The surface deformation responds to the small (<20 MPa) temporal changes of the stress field associated with the subduction earthquake cycle. This transient stress is superimposed on the larger background regional stress field in which the maximum compression is parallel to the margin. The weakness of the Cascadia subduction thrust fault and the unusual stress state of the forearc region have important implications for earthquake hazards. For example, a subduction earthquake may induce large strike‐slip earthquakes in the forearc that affect a large area.

Continuous GPS monitoring of elastic strain in the Northern Cascadia Subduction Zone

Previous monitoring and modeling of crustal deformation across the northern Cascadia margin at Vancouver Is. has provided strong evidence that the subduction thrust fault is locked and may generate future great earthquakes. The recent establishment of the Western Canada Deformation Array (WCDA), a network of continuous GPS trackers in southwestern British Columbia, provides a new tool for monitoring crustal strain and thereby helps to assess the earthquake hazard in this region. 17 months of continuous data from the three longest running WCDA sites (Penticton, Victoria, Holberg) indicate: 1.) A 7 mm/yr easterly motion of Victoria relative to Penticton. Victoria is located in the forearc at the southeastern end of Vancouver Is., 230 km from trench axis, whereas Penticton is located behind the arc, 520 km from the trench axis, and assumed fixed on the stable North American plate; 2.) A 3 mm/yr northwesterly motion and 10 mm/yr uplift (with a large uncertainty) of Holberg relative to Penticton. Holberg is located on the northern‐most part of Vancouver Is., 50 km from the trench and near the plate triple junction; and 3.) The presence of probably seasonal variations in the apparent relative positions of the two coastal sites with respect to Penticton. The landward motion of Victoria agrees with the deformation predicted by elastic dislocation models of the interseismic period for a great subduction‐thrust earthquake as constrained by previous deformation data. Although not as well resolved, the motion of Holberg, roughly parallel to the margin, is significantly different and is inconsistent with simple subduction models. The non‐linear variations in measured position are dominated by an annual period probably resulting from seasonal biases in the precise orbit estimates or in the modeling of tropospheric delays or tidal effects.

Upper mantle structure of the northern Cascadia subduction zone

Previous knowledge of the structure of the Cascadia subduction zone north of the Canada–United States border has been derived from a variety of geophysical studies that accurately delineated the downgoing Juan de Fuca plate from the offshore deformation front to depths of ~50–60 km beneath south-central Vancouver Island and the Georgia Strait. Little is known, however, of the structure of the Cascadia subduction zone farther westward and to greater depths in the upper mantle. We have assembled a set of some 1100 teleseismic traveltimes from events recorded on the Western Canadian Telemetered Network to augment a previously existing data set recorded on the Washington Regional Seismograph Network. The composite data set is inverted for upper mantle structure below Washington, Oregon, and southwestern British Columbia. We analyze the new northern portion of the model between 48.5–50°N and 118–127°W, which provides the first images of the deep slab structure in this region. The model is parameterized using splines under tension over a dense grid of knots. The nonlinearity of the inverse problem is treated by iteratively performing three-dimensional ray tracing and linear inversion. Resolution tests performed with a synthetic slab model indicate that the deep structure is resolved by the data north to at least 50°N. The inversions are characterized by a quasi-planar, high-velocity body inferred to represent the thermal and compositional anomaly of the subducted Juan de Fuca plate. This body exhibits velocity deviations of up to 3% from the background reference model and extends to depths of at least 400–500 km. The depth contours of the slab in the upper mantle mimic those of the shallow slab by changing strike, in the latitude range 48.0–48.5°N, from north–south in Washington to northwest–southeast in southern British Columbia. This forces the development of two arch-type structures: a main arch observed in previous studies trending east–west over Puget Sound and a possible second arch extending northeasterly from the Georgia Strait into the British Columbia interior. A steepening of the deep slab dip from British Columbia south towards Puget Sound and complexity in the evolution of the arches in depth may be the result of a change in plate motions at 3.5 Ma associated with the detachment of the Explorer plate.

Finite element study of uplift and strain across Vancouver Island

Geological evidence for sudden coastal subsidence along the west coast of southern Vancouver Island points to the occurrence of great prehistorical subduction earthquakes. Contemporary uplift and crustal shortening patterns in southern Vancouver Island appear to indicate that the subduction megathrust fault is currently locked. To understand better the dynamics of the observed surface deformation, we develop a finite element model of earthquake cycles for the northern Cascadia subduction zone across southern Vancouver Island, using a linear viscoelastic rheology. The model consists of the continental and oceanic lithospheres, the asthenospheric mantle with a viscosity of 5 × 1019 Pa∙s, and a low-viscosity (1018 Pa∙s) mantle wedge between the subducted oceanic plate and the overlying continental plate. The shallow geometry of the subducted Juan de Fuca plate is well defined by the results of various geophysical surveys, and the deep geometry is constrained by the results of seismic tomography. The model megathrust fault has a stick-slip zone near the surface, a viscoelastically weakly coupled zone (viscosity 7 × 1017 Pa∙s) at depth, and a narrow free-slip zone in between. Earthquakes are allowed to occur every 500 years. Varying the recurrence time does not greatly affect the surface deformation in the later part of the interseismic period. Experiments varying the width of the stick-slip zone lead to the conclusion that a width of about 70 km satisfies both the observed coseismic coastal subsidence and the contemporary surface deformation pattern. The results of a simple elastic dislocation model for thrust earthquakes that had been previously applied to the region are compared with the solutions of the viscoelastic model. Despite its simplicity, the elastic model approximates well the surface deformation of the viscoelastic model in the second half of the interseismic period, although it predicts a slightly narrower stick-slip zone of the fault. The present viscoelastic model is limited principally by the two-dimensional approach, the assumptions of purely stick-slip behaviour of the thrust fault, and the uncertainties in rock rheology.

Current deformation and the width of the seismogenic zone of the northern Cascadia subduction thrust

Evidence has been obtained for the accumulation of elastic strain across the northern Cascadia subduction zone that may be released in a future very large subduction thrust earthquake. Vertical and horizontal strain rates across the southern Vancouver Island region have been determined through (1) long‐term trends in tide gauge data, (2) changes in repeated accurate leveling surveys, (3) changes in repeated high‐accuracy gravity profiles, and (4) horizontal shortening observed in repeated precise positioning surveys. The outer coast is uplifting at a rate of a few millimeters per year decreasing landward, and shortening is occurring across the 100‐km‐wide coastal region at a rate of about 0.1 microstrain per year (mm km−1yr−1). The results are compared with the distribution of strain accumulation predicted from elastic dislocation and viscoelastic models for a subduction thrust fault. The location of the fault as used in the models is well defined by multichannel seismic reflection and other geophysical data. Most of the observed current deformation can be explained by interseismic strain accumulation associated with the subduction thrust of southern Vancouver Island and northern Washington, provided the locked portion is restricted to a 60‐km‐wide band offshore beneath the continental shelf and slope. This conclusion also results from modeling the coseismic subsidence on the outer coast of Vancouver Island about 300 years ago deduced from paleoseismicity data. The unusually narrow downdip extent of the subduction thrust seismogenic zone, that extends little if at all beneath the coast, is a consequence of high temperatures associated with the young age of the subducted oceanic lithosphere and the thick blanket of insulating sediments. The high temperatures limit brittle seismogenic behavior downdip to where the thrust fault is at a depth of less than 15 km. The distance from the seismic portion of the megathrust limits the estimated ground motion at the major centers of Vancouver and Victoria from this source. The narrow width may also limit the earthquake size; however, events of magnitude well over 8 are possible.

Tectonic sediment thickening, fluid expulsion, and the thermal regime of subduction zone accretionary prisms: The Cascadia Margin off Vancouver Island

Tectonic sediment thickening, fluid expulsion, and the resulting thermal regime of the northern Cascadia subduction zone accretionary prism have been described by numerical models. Both the sediment thickening and the fluid expulsion are found to have important thermal effects. Constraints on the models are provided by (1) detailed heat flow profiles across the continental slope accretion zone from variations in the thermally controlled depth to a gas hydrate bottom‐simulating reflector (BSR), and (2) porosity changes across the accretion zone inferred from the landward variation in the P wave velocity‐depth profiles. The heat flow in the region of rapid prism thickening centered 15 km landward of the deformation front is 20% below that predicted by a larger scale thermal model that ignores the thermal effects of the sediment accretion processes. The porosity also is substantially higher in this region compared to at the same depths in the adjacent deep sea Cascadia basin. The model results are in agreement with both the thermal and velocity data if the incoming sediment section is initially simply tectonically shortened horizontally and thickened vertically, generating a high porosity underconsolidated section. Reconsolidation and fluid expulsion in response to the tectonic thickening is slow, occurring mainly 20–30 km landward of the deformation front. The maximum predicted fluid expulsion rate is about 4 × 10−11 m s−1 (1.3 mm a−1). The pore pressure distribution across the deformation front region has been estimated from the velocity data using the simplifying assumption that velocity is only a function of effective pressure; pore pressures are high, reaching 80% of lithostatic 15 to 20 km landward of the deformation front. An estimate of the bulk permeability in the seaward tens of kilometers of the prism ranging from 10−16 m2 near the top to about 10−18 m2 near the base has been obtained through matching the pore pressure distribution in the numerical model to that obtained from the velocity data. The prism sediments are inferred to be substantially underconsolidated and pore pressures to be high for at least the seaward 30 km of the accretionary prism.

S wave velocity structure of the Northern Cascadia Subduction Zone

Teleseismic receiver functions from an array of portable broadband seismograph stations located in southwestern British Columbia are interpreted to estimate the S wave velocity structure to upper mantle depths across the northern Cascadia subduction zone. At our westernmost station on central Vancouver Island, a prominent low‐velocity zone (ΔVs = −1 km s−1) with a high Poisson's ratio is estimated at 36–41 km depth. This feature correlates with the reflective “E” zone observed in LITHOPROBE reflection data, a region that also exhibits high electrical conductivity and has recently been interpreted as a fluid‐saturated shear zone above the subducting Juan de Fuca (JdF) plate. Analysis of data at our stations to the NE permits this zone to be mapped into the upper mantle, 54 km beneath the British Columbia mainland, and approximately 250 km from the locus of subduction. The subducting oceanic crust is imaged at 47–53 km depth dipping 15°± 5° in the direction N30°E ± 20° beneath central Vancouver Island. The dip angle increases to 22° ± 5° at a depth of 60–65 km beneath the Strait of Georgia. The results of this analysis provide the first definitive evidence for the location of the subducting plate in this region and indicate that the seismicity at depth occurs within the oceanic crust. The dip direction of N30°E at each station provides new evidence that the JdF plate is arched upward as it subducts in this region. A low‐velocity zone that coincides with the “C” reflectors beneath Vancouver Island is interpreted to extend east to the British Columbia mainland near the base of the North American crust. The top of this zone, near 20–26 km depth, lies near the lower limit of shallow seismicity. The continental Moho is estimated to be 36 km below the Strait of Georgia.

Comparison of dynamic source parameters for earthquakes in different tectonic regions of the northern Cascadia subduction zone

Based on characteristics of displacement amplitude spectra of near-distant S-waves from 52 earthquakes in the magnitude ( M L ) range −0.6 to 5.4, dynamic source parameters were computed and compared for three tectonic regions in the northern Cascadia subduction zone: a downgoing slab of the Juan de Fuca Plate with earthquake depths of about 30 km; the same slab further east with deeper events, about 60 km; and an overriding part of the North American Plate. No distance dependence of observed corner frequencies was obtained, nor any significant regional difference in source parameters, although there is an indication of smaller stress drop in the North American Plate for larger events. Factors which may distort a regional comparison were examined, but neither source radiation directivity nor propagation path and station site characteristics have significant influence. The structural model has influence, however, and two different density/velocity models were used for the deeper Juan de Fuca slab events. Seismic moments (M 0) range from E + 17 to 0.4 E + 24 dyn cm, source radii from 0.07 to 1.6 km, stress drops from less than 0.1 to 40 bar and average dislocations from less than 0.1 to 7 cm. In the range E + 18 to E + 21 dyn cm, in which the vast majority of the events are confined, both the stress drop and displacement increase rapidly with event size (seismic moment or magnitude) as the seismic moment increases rapidly with fault dimension. The relation between seismic moment (dyn cm) and M L magnitude computed from data from all three regions is similar to relations presented for California and reads log( M 0) = 17.65 + 1.02 M L . There are, however, differences between events in the North American Plate and the Juan de Fuca Plate, with larger M L for the former at equivalent seismic moment.

A seismic study of methane hydrate marine bottom simulating reflectors

Multichannel seismic reflection data have been analyzed from an area of clear bottom simulating reflectors (BSRs) on the northern Cascadia subduction zone margin off Vancouver Island. The reflector at a depth of about 300 m subbottom is interpreted to represent the base of a layer of methane hydrate or clathrate. The shallow water depth of 1300 m and the 3600‐m‐long hydrophone array have allowed BSR amplitude‐versus‐offset and high‐resolution velocity analysis, as well as modelling of vertical incidence data. The results of all three types of analysis can be best explained by a 10 to 30‐m‐thick high‐velocity layer located immediately above the BSR about 300 m below the seafloor, having a sharp base and transitional top. In the layer, about one third of the sediment pore spaces must be filled with hydrate “ice”. There is no seismically detectable free gas beneath the BSRs. These results put important constraints on models for the distribution and formation of BSR hydrate.

Shear wave constraints on a deep crustal reflective zone beneath Vancouver Island

Analysis of teleseismic receiver functions recorded in the northern Cascadia subduction zone show that large P‐to‐S converted phases are generated at the boundaries of a dipping layer of deep crustal reflectors (the “E” zone) beneath central Vancouver Island. These data provide the first shear wave constraints on this layer which is clearly observed in seismic reflection sections and is coincident with a region of low electrical resistivity interpreted from a magnetotelluric survey. Modeling of the amplitudes, polarities, and arrival times of the P‐to‐S converted phases demonstrates that the E zone is a prominent low‐velocity layer, with an S velocity 1.0±0.2 km/s lower than the regions above and below. The E zone dips 7°±5° in the direction N10°E±20° and extends from depths of 36.5 to 41 km at ALB‐B on central Vancouver Island. The estimated S velocity contrast, when considered with refraction P velocities, yields a Poisson's ratio of 0.34 for this zone. Given the uncertainty in ΔVp and ΔVs, this value may range from 0.27 to 0.38. The low P and S velocities and high Poisson's ratio combined with electrical resistivity measurements suggest a porosity of 0.1–1.9% and crack aspect ratios of 0.001–0.01, values consistent with a region containing thin, fluid‐saturated cracks and thus supportive of the interpretation of this feature as a major shear zone.

Water budgets in accretionary wedges: a comparison

Direct or indirect measurements of fluid flow out of the toe of accretionary wedges have now been made in the Barbados, Central Oregon, Northern Cascadia and Nankai subduction zones. The steady-state local compaction model predicts velocities of fractions of a millimetre per year and total outflows from the toe of a few cubic metres per year per metre of length along strike of the subduction zone (m2 a-1). Sea bottom measurements reveal channellized flows at velocities of hundreds of metres per year and total outflows from the toes of a few hundreds of square metres per year. Thermal arguments show in the Nankai area that all of this large surface flow cannot come from as deep as the decollement and that consequently a significant dilution by shallow sea water convection must be present. We propose that this convection is driven by the reduced density of less saline fluid of deep origin. Thus the outflow of water of deep origin may be only a few tens of square metres per year. We note, however, that there are indications in the Barbados area of massive flow of low salinity water at depth along the decollement. These flows imply the existence of large scale non-steady-state lateral transport and require the existence of sources of low salinity water. These might include the smectite-illite transformation and the fluid contained in the oceanic crust. However, these sources are limited and the possibility exist that other more important sources may be required such as long distance transport of fresh water from adjacent sedimentary basins and (or) recharge mechanisms such as the seismic pumping one.

Rates of fluid expulsion across the Northern Cascadia Accretionary Prism: Constraints from new heat row and multichannel seismic reflection data

One hundred and ten closely spaced probe heat flow measurements provide new constraints on the thermal regime of the northern Cascadia accretionary sedimentary prism off Vancouver Island. Complementary heat flow values have been obtained from the depth of a bottom‐simulating seismic reflector (BSR) that is interpreted to mark the thermally controlled base of a methane hydrate layer. The only local heat flow variations observed are associated with a sediment slump that is seen in SeaMARC II acoustic images and with the outcrop of several major thrust faults. These anomalies appear to be the result of material displacement only; strong localized fluid flow along faults or elswhere is not evident in the probe or BSR results. Fluid expulsion resulting from the dewatering of the prism sediments appears to occur regionally in the 10–20‐km‐wide zone landward of the deformation front. In this area there is a significant disagreement between the probe and BSR heat flow estimates (roughly 30%) that can be explained by a regionally uniform vertical fluid flow at a rate of about 8×10−10 m s−1. This is in good agreement with the estimated fluid expulsion rate required by the decrease in porosity landward of the deformation front, as estimated from the increase in seismic velocities derived from multichannel reflection data. The heat flow in Cascadia Basin seaward of the deformation front, corrected for the effect of sedimentation, is in excellent agreement with that predicted by cooling plate models. Landward, there is a regional trend of decreasing heat flow across the accretionary prism, which is consistent with a model of simple tectonic thickening. Temperatures at the top of the oceanic crust are estimated to be over 200°C beneath the 2–3 km of sediment in Cascadia Basin. Temperatures at the interface between the prism and the oceanic crust continue to increase landward, and reach 400°–450°C beneath the middle to inner continental shelf. Initiation of megathrust earthquake failure along the main subduction thrust may be thus restricted by the high temperatures to the zone beneath the continental slope and outer shelf. Early and shallow metamorphism of the prism material probably is another important consequence of the early deep burial of the young oceanic crust in this setting.

The northern Cascadia subduction zone at Vancouver Island: seismic structure and tectonic history

The structure and Tertiary tectonic history of the northern Cascadia subduction zone have been delineated by a series of new multichannel seismic lines acquired across the continental shelf to the deep sea, combined with adjacent land multichannel seismic data and results from a wide range of other geophysical and geological studies. The top of the downgoing oceanic crust is imaged for a remarkable distance downdip from the deep ocean basin to a depth of 40 km beneath Vancouver Island. The reflection depths are in good agreement with seismic refraction models and Benioff–Wadati seismicity. Two broad reflective bands imaged as dipping gently landward at depths of about 15 and 30 km on the land lines merge to a single reflector band offshore. They may represent underplated oceanic material or, alternatively, they may not be structural but may be zones of contrasting physical properties, perhaps representing trapped fluid. Two narrow terranes, the Mesozoic marine sedimentary Pacific Rim Terrane and the Eocene marine volcanic Crescent Terrane, have been thrust beneath, and accreted to, the margin in the Eocene, about 42 Ma, near the start of the present phase of subduction. They provide a landward-dipping backstop to the large sediment wedge accreted since that time. The deformation front is characterized by mainly landward-dipping thrust faults that cut close to basement. This result and the mass balance of the incoming sediment compared with that present in the accreted wedge suggest that there is little subduction of sediment into the mantle. The Tofino Basin sediments, up to 4 km in thickness, have been deposited on the continental shelf over the accreted terranes and the developing accretionary wedge.

Accretion and recent deformation of sediments along the northern Cascadia subduction zone

New multichannel seismic reflection profiles, acoustic imagery, and swath bathymetric data from the Vancouver Island continental margin reveal the surface morphology, cross-sectional geometry, and deformation style of this part of the Cascadia accretionary prism. The incoming 3- to 5-km-thick sediment section fails most commonly along landward-dipping thrusts that penetrate to or very close to the top of the underlying oceanic crust. The faults are listric with approximately circular cross sections, and hanging-wall sediments are back-tilted in a planar manner. High, near-lithostatic pore pressures in the deeper part of the section are inferred from the flattening of the faults near the decollement, and from the long distances over which the decollement has propagated seaward. The faults dip more steeply at 40°-45° near the surface where Pleistocene turbidites are displaced. Similar folds have developed where the turbidites are particularly thick (>4 km), and near-hydrostatic pore pressures are inferred to occur in the turbidite part of the section. An increase in pore pressure with depth is expected as a consequence of sedimentation and tectonic loading, particularly because the deeper hemipelagic sediments are probably far less permeable than are the overlying turbidites. The top of the downgoing Juan de Fuca plate, which defines the base of the accreted sedimentary wedge beneath the margin, is well imaged in the new seismic reflection profiles and is further defined by land reflection data, seismic refraction, and earthquake hypocenters. The landward-dipping continental backstop to the prism is also well located. If these constraints are used for the geometry of the accretionary prism, a general balance is found between the present prism volume and the estimated total sediment supply over the approximately 42-m.y. accretion history. This result, along with the observation that deformation-front faults penetrate most or all of the sediment section, suggests that little or no sediment has been subducted since initiation of the present phase of subduction in the Eocene. Application of the critically tapered wedge model of Davis and others to the observed prism geometry suggests that high average pore pressures occur in the prism, particularly the seawardmost part beneath the lower slope. Beneath the continental shelf, the dip of the subducting slab increases to greater than 11°, a dip that according to the critically tapered wedge model, is too steep to allow upward growth of an accretionary prism. This provides a simple explanation for ongoing subsidence of the Tofino sedimentary basin over the older part of the accretionary complex. To the southeast, the Juan de Fuca plate dips much more gently beneath the Olympic Peninsula of northern Washington, allowing the accretionary prism to grow upward well above sea level to form the Olympic Mountains.

Dipping Seismic Reflectors, Electrically Conductive Zones, and Trapped Water in the Crust Over a Subducting Plate

The LITHOPROBE program across the northern Cascadia subduction zone at Vancouver Island has obtained geophysical data that suggest a dipping zone of trapped free pore water in the overlying continental crust. Land and marine multichannel seismic reflection profiles show several bands of reflectors. The primary one at a depth of about 30 km is about 5 km thick, dips inland, and appears to coincide with an electrically conductive layer defined by magnetotelluric data. The downgoing oceanic crust is imaged as a thin reflector dipping at about 15° and 10 km deeper than the primary reflector, a location that is consistent with seismicity and seismic refraction data. Both the reflectors and the conductor have been modeled by near‐horizontal lamellae with 1–4% saline water porosity. Detailed geothermal measurements defining heat flow that decreases inland indicate that the dipping bands are nearly isothermal at a temperature of 400°–500°C. It is suggested that free water generated by dehydration of the downgoing oceanic plate migrates upward until reduced temperature results in hydration reactions and mineral precipitation that forms an impermeable barrier. Low vertical permeability is required to keep the water from separating to the top of the layer. This can be reconciled with the low electrical resistivity if pore interconnection is through thin grain boundary tubes as predicted by theoretical equilibrium pore geometry, although very small effective grain size is necessary. An abrupt transition from high permeability to very low permeability is predicted at a several percent porosity, in agreement with the modeled porosity. Such pore geometry with low permeability may be restricted to temperatures above about 400°C, i.e., in greenschist to amphibolite facies conditions, where grains will deform to equilibrium geometry and cracks are annealed. The region above the reflective and conductive band is estimated to be within blueschist conditions, consistent with the inferred high seismic velocity from refraction data and high density from gravity data.

Control of Clay-Mineral Stratigraphy by Selective Transport in Late Pleistocene-Holocene Sediments of Northern Cascadia Basin--Juan De Fuca Abyssal Plain: Implications for Studies of Clay-Mineral Provenance

ABSTRACT Previous investigations (Duncan and others, 1970; Stokke and Carson, 1974) of Cascadia Basin clay mineralogy indicate a general trend in which montmorillonoids (smectites, 17 A) are enriched relative to illite (10 A) and chlorite (7 A) in Holocene lutites, whereas illite and chlorite dominate in Late Pleistocene deposits. The apparent of all clay-mineral groups declines from Pleistocene to Holocene (Russell, 1967). There exist, however, numerous local exceptions to these general patterns which are not readily explained by the previous works. Fifteen samples from 5 cores in Northern Cascadia Basin and Juan de Fuca Abyssal Plain were fractionated into 6 size classes. Semi-quantitative determinations of clay mineralogy and crystallinity were made on each of the 90 subsamples. The results indicate a definite size-dependency in clay mineralogy and which is consistent both areally and temporally. Discriminant-function analysis indicates that clay mineralogy and of some individual size fractions are statistically distinct but that the stratigraphic units (Late Holocene, Early Holocene, and Late Pleistocene) are not statistically discernible. Textural variations within and between Holocene and Pleistocene parallel sediment clay-mineral changes. and mineralogy can be predicted on the basis of clay-min ral size-dependency and grain-size distribution. It is inferred that the stratigraphic textural-mineralogic variations result from selective transport and deposition and not from contrasting source areas or weathering conditions, as had previously been suggested (Duncan and others. 1970; Stokke and Carson. 1974). Future investigations of clay-mineral provenance, dispersal, or correlation, in any area, should be based upon data which has been corrected for the effects of clay-mineral size-dependency.